• Open-ecosystem AI strategy: AMD is driving a three-pillar AI approach centered on reliable annual leadership product roadmaps, deep partnerships with open-source software communities, and co-development with hyperscalers like Meta and OpenAI.
• Inference efficiency curve: AI workloads are achieving roughly 10x generational efficiency gains per year in tokens delivered per dollar and megawatt, positioning inference as the primary driver of widespread deployment and monetization.
• Agentic compute demand: The rise of autonomous agent-to-agent workflows is shifting compute requirements, significantly increasing demand for CPUs to handle orchestration, planning, and task execution alongside GPUs.
• India engineering footprint: AMD employs over 10,000 people in India, with local engineering contributing to virtually every product portfolio, while the company explores expanding regional partnerships and data center opportunities.